The aims of this study are to determine the level of activity of the hemostatic system (in vivo thrombin generation) in patients who have suffered a previous ischemic stroke. We will measure the plasma level of the prothrombin activation fragment Fl+2, in a representative subset of patients in the Warfarin Aspirin Recurrent Stroke Study (WARSS - NINDS 5 R0l NS2837I). In that study, patients who have suffered a "cryptogenic" (embolic) stroke are assigned to either antithrombotic or antiplatelet therapy. The treatment arms consist of warfarin to maintain INR between 1.4 and 2.8, or aspirin at 325 mg/day. For this substudy of WARSS, we hypothesize that Patients with previous embolic stroke will have a higher level of Fl+2 than patients with other types of ischemic stroke. We will determine whether prothrombin fragment Fl+2 is lowered to a greater extent by warfarin than by aspirin in the embolic "cryptogenic" group of patients as compared to patients in the other stroke groups (atherothrombotic, tandem arterial pathology, small vessel/lacunar). In addition to prothrombin fragment Fl+2 and INR, we will measure serum fibrinogen, albumin/globulin ratios and complete blood counts. To determine whether the relationships between the various parameters are stable over time, plasma samples will be obtained every six months for 2.5 years. Our proposed studies provide the opportunity to answer the following questions: (a) is there a particular group of patients that have a higher level of prothrombin generation?; (b) is there a particular subgroup of patients who are resistant to warfarin-induced thrombin generation?; (c) if warfarin is effective in preventing secondary stroke at the target INR, what is the relation between INR and Fl+2?; and (d) are these relationships stable over time?